Age-Associated Differences in Paddock Locomotor Activity Among Senior Horses: A Pilot Observational Study

Turnout locomotor activity is a potentially informative indicator of health and welfare in older horses, yet objective field data in seniors remain limited. We examined whether a brief turnout recording could detect cross-sectional associations between chronological age and locomotor activity in senior horses in this study setting. In this single-site observational study, 28 senior Selle Français horses (17-35 years) contributed 122 paddock sessions (2 h each), with total distance and mean speed quantified using a Polar Team Pro sensor. Associations with age were assessed using linear mixed-effects models adjusted for temperature and precipitation. Age was decomposed into between-horse and within-horse components to separate cross-sectional from within-horse information. Raw (untransformed) total distance ranged from 148 m to 3994 m (median 1128 m; IQR 638-1779 m; mean 1292 ± 834 m). Log-transformed total distance was negatively associated with age (β = -0.062 per year; 95% CI -0.094 to -0.032; p < 0.001), driven by the between-horse component (β = -0.063; q = 0.003). The within-horse estimate was imprecise and not statistically supported (p = 0.75). Mean speed showed a similar pattern, with a significant between-horse association (β = -0.060; q = 0.003) and an imprecise within-horse estimate (p = 0.87). These findings suggest that brief paddock actimetry may help characterize between-horse heterogeneity and support group-level welfare monitoring. However, the present dataset does not allow robust inference about within-horse ageing trajectories or individual-level biological ageing. Larger multi-site cohorts with denser follow-up and external validation are needed before individual trajectories or clinical interpretation can be established.